Solving a Second Temple Puzzle

Frankie Snyder with proposed reconstructions of floor tiles from Second Temple

Frankie Snyder is not an archeologist. “They won’t let me use the title,” she says of her coworkers. “I’m a mathematician–I don’t have an archeology degree.” But despite not being an archeologist, she may have figured out the design of the decorative floor tiles in the Second Temple. I spoke to Frankie at the Annual Archaeological Conference in Ir David, the City of David. This year’s conference was titled “Digging for Truth.”

Frankie was publicly showing her reconstructions of the Temple floor designs for the first time. Several of the designs had been published in newspaper articles earlier in the week. She was also scheduled as a speaker during the oral presentations later in the evening. It is no doubt unusual to feature a mathematician at an archeological conference. Despite her lack of credentials, many of the thousand plus attendees at the conference stopped to view display.

Examples of several opus sectile reconstructed designs. Small black bitumen triangle on tile at top of picture was recently found and will put in place in the reconstruction.

The Temple Mount where the Al Aksa Mosque and Dome of the Rock stand today is administered by an Islamic trust, the Waqf. From 1996 to 1999 the Waqf constructed an underground mosque on the southeast corner of the mount. They did not perform the customary careful survey to make sure all historical artifacts were found and preserved. Instead, they excavated with bulldozers and removed large shovels full of earth at one bite. About four hundred truckloads of soil,  were dumped in the Valley of Kidron. Uncounted numbers of artifacts were unceremoniously removed from the site.

Many Israelis were upset by this wholesale destruction of a historical and religious site. The government decided not to interfere in the Waqf’s jurisdiction. The Waqf completed its work.

When archeologists discover artifacts, they dig  carefully and take note of the exact surroundings. Examination of items in situ is a major factor in determining their age and historical period. Because the remains are hundreds of years old, care must also be taken to make sure removal does not damage them. Using heavy equipment to excavate destroys both the physical situation and often the integrity of the objects.

Archeologist Dr. Gabi Barkay felt it was crucial to recover and preserve artifacts removed from the Temple Mount. In 2004, he and Zachi Dvira initiated the Temple Mount Sifting Project. The purpose of the unique project is to find, analyze, and identify artifacts in the debris. They use a wet sifting technique they developed specifically for the task. Participants in the project wash the dirt-encrusted debris, sifting out anything that shows traces of human work: bits of pottery, stone with flat edges, metals, coins, jewelry, and other things. The public is invited to participate in two hour long programs. After they are introduced to the project, they sift several buckets of debris. Several times a year newspapers publish photos of smiling tourists holding some remarkable find.

Just after she made aliyah from the U.S. in 2007, Frankie decided to visit the Sifting Project for a day. She’s still there, although now as a paid employee.

During the course of her work, she came across fragments of opus sectile tiles. These are colored stones cut specifically to fit into a design. Unlike mosaics, opus sectile uses no mortar to hold the design in place. The stones are cut to fit exactly, so tight you can’t fit even a sharp blade between them. Because the stones must be cut so precisely, opus sectile is considered more elegant and prestigious than mosaics.

Opus sectile was developed in Rome and brought to the Middle East by Herod. All of his palaces contain this type of floor, frequently in the bathhouse. Some of the Roman period mansions in Jerusalem’s old city have this type of floor in a room or two. The Byzantines, Crusaders, and later Muslims also used this technique to decorate their impressive buildings.

As Frankie worked at the sifting project, she gradually became an expert on opus sectile tiles. One of the major problems she faced was sorting the tiles. Because the soil had been removed from the Temple Mount in large mechanical shovels, debris from many periods was jumbled together. So one of her first tasks was to learn to differentiate Herodian tiles from tiles used in a later period.

Samples of types of rock used in Herodian tiles of Temple floor

The sorting is based on four criteria: the type of stone used, the basic geometry of the piece, comparison with pieces from other sites whose age is known, and from historical sources. She also looked at the color of the stones and the craftsmanship with with they had been cut. As she started to develop her expertise, she was asked by other archaeologist to visit sites at which they were working. Examining tiles from sites known to have been built by Herod, such as Masada, Jericho, and Kypros, further advanced her knowledge. Additionally she worked with some Italian experts on opus sectile floors, including Lorenzo Lazzarini, who has developed methods of identifying the quarry of origin of ancient marble tiles.

The first place she reconstructed a floor design from fragments was Banyas. They had 172 pieces from a floor that originally consisted of 25,000 pieces of stone. She was already familiar with Roman tile designs that were common in Israel: the strips of triangles, the pinwheel, the four- and eight-pointed stars, the three square-within-square design, and the four square-in square pattern. She knew that each design measured one Roman foot, about 29.6 cm. (11.60.5 inches).

“It sounds like putting together a jigsaw puzzle, but the box contains pieces from half a dozen different puzzles, and you don’t have all the pieces.” I said.

“Except you don’t have a picture on the box lid, and some of the pieces are broken,” Frankie added.

The Temple floor was much harder than Banyas, because the floor was much larger. Out of an estimated million pieces, they had found less than three hundred. The Herodian designs often deviated from the Roman patterns. For example, the triangles design utilized isosceles triangles, whose base was the same length as its height. The three angles were 64, 64, and 52 degrees. One corner fragment could determine the design. The eight pointed star also has unique angles in its center, so the small piece of pink limestone with that specific angle would have been part of an eight-pointed star design.

After determining the design a specific piece fit in, Frankie made a reconstruction of the design. Most of the reconstructed tiles contain only two to four stones or stone fragments. She pasted paper of matching color on stone to show how the whole tile would have looked. Three of the pink corners of one tile were photocopies of the actual pink stone in the other corner.

Frankie Snyder showing cut edge of a stone used in opus sectile floor tiles

It was amazing to see how tightly some of the stones in the reconstructions fit together. The Herodian workmen did not have the precision tools we have today. She explained that the stones were cut to size using a saw, sand, and water. The actual cutting was done by the sand; the saw forced down into the line of the cut. I felt the side of the stone she held out to me. I was amazed at how smooth it felt.

When someone asked if the reconstructions had been difficult, Frankie shook her head. “It’s just simple plane geometry.”

Simple or not, it was awesome to realize the small bitumen triangle I held in my hand had once been part of the floor of the Temple. These tiles would have been in a covered area of the Temple complex, in the stoa or portico. The work that went into the these floor tiles was too valuable to leave exposed to the elements.

Others were just as affected by these reconstructions as I was. One woman bent down to the table on which the tiles sat and kissed the original stone fragment one of them contained. Another person picked up a tile, turned to face north, towards the Temple Mount, and said a blessing.

Frankie illustrated her short lecture with slides of her work. Despite speaking in English, she received more applause for her presentation than any other speaker. I suspect she would have received that applause even if she had stood there silently while simply showing the slides. Her work speaks for itself.